Daily Papers of 2024-12-23

  1. Parallelized Autoregressive Visual Generation 47 upvotes, #1 of 2024-12-23
  2. Offline Reinforcement Learning for LLM Multi-Step Reasoning 33 upvotes, #2 of 2024-12-23
  3. SCOPE: Optimizing Key-Value Cache Compression in Long-context Generation 18 upvotes, #3 of 2024-12-23
  4. CLEAR: Conv-Like Linearization Revs Pre-Trained Diffusion Transformers Up 18 upvotes, #3 of 2024-12-23
  5. Taming Multimodal Joint Training for High-Quality Video-to-Audio Synthesis 16 upvotes, #5 of 2024-12-23
  6. Toward Robust Hyper-Detailed Image Captioning: A Multiagent Approach and Dual Evaluation Metrics for Factuality and Coverage 13 upvotes, #6 of 2024-12-23
  7. MixLLM: LLM Quantization with Global Mixed-precision between Output-features and Highly-efficient System Design 12 upvotes, #7 of 2024-12-23
  8. TRecViT: A Recurrent Video Transformer 11 upvotes, #8 of 2024-12-23
  9. Sequence Matters: Harnessing Video Models in 3D Super-Resolution 10 upvotes, #9 of 2024-12-23
  10. IDOL: Instant Photorealistic 3D Human Creation from a Single Image 5 upvotes, #10 of 2024-12-23
  11. Multi-LLM Text Summarization 5 upvotes, #10 of 2024-12-23
  12. LLMs Lost in Translation: M-ALERT uncovers Cross-Linguistic Safety Gaps 4 upvotes, #12 of 2024-12-23
  13. Fietje: An open, efficient LLM for Dutch 3 upvotes, #13 of 2024-12-23

Data: hysts-bot-data/daily-papers-stats and the Daily Papers API. Open data: tardellirs/paper-pulse-data. Sister project: Model Pulse, the download history of every model on the Hub.